AI labs are buying and pulping rare books to scan them for training data
xuanalogue · x · 2026-07-27
The post highlights reporting that AI labs are buying books by the pallet, cutting spines off, scanning them, and pulping the originals. It says the ordering system is keyed to ISBNs and does not account for rarity, so some potentially last-copy books may be destroyed.
It also notes the copyright/fair-use angle behind the pipeline: a federal judge reportedly ruled the practice fair use because only one copy exists at a time after the originals are destroyed. The thread cites a report that Anthropic hired a former Google Books partnerships head to help obtain “all the books in the world.”
More from Companies & People
- Enterprise teams are still spending four years on Java upgrades because current models can’t handle them — curious_vii · 2026-07-28
- Google DeepMind reconstructs Pelé’s lost 1959 goal and opens GNM to developers — plopesresearch · 2026-07-28
- Thinking Machines appears as a 2025 seed-backed private startup led by Mira Murati — suchenzang · 2026-07-28
- LangChain says open source is in its blood as it partners with NVIDIA — Hacubu · 2026-07-28
- Mistral opens registration for its Brooklyn cat café coworking event — sophiamyang · 2026-07-28
- A Fields Medalist joins OpenAI as $250B data-center financing and Kimi K3 land — Dapper-Tale-4021 · 2026-07-28